EJB performance issues migrating from JBoss 4.0.0 to JBoss EAP 4.3.0

Solution Verified - Updated -

Issue

  • The same load test completes fine on JBoss 4.0.0 but fails on JBoss Enterprise Application Platform (EAP) 4.3.
  • EJB pool sizes not evenly distributed amongst 4 nodes. For 45 concurrent users, our pool sizes grew to 37, 20, 16 and 12.
  • The following OutOfMemoryError occurs on the node with the highest pool size (not always the same node):

    11:54:04,738 ERROR [UDP] failed unmarshalling message
    java.lang.OutOfMemoryError: GC overhead limit exceeded
            at org.jgroups.util.Util.readAddress(Util.java:569)
            at org.jgroups.ViewId.readFrom(ViewId.java:134)
            at org.jgroups.View.readFrom(View.java:213)
            at org.jgroups.util.Util.readStreamable(Util.java:692)
            at org.jgroups.protocols.pbcast.GMS$GmsHeader.readFrom(GMS.java:1129)
            at org.jgroups.Message.readHeader(Message.java:711)
            at org.jgroups.Message.readFrom(Message.java:634)
            at org.jgroups.protocols.TP.bufferToMessage(TP.java:977)
            at org.jgroups.protocols.TP.handleIncomingPacket(TP.java:833)
            at org.jgroups.protocols.TP.access$400(TP.java:47)
            at org.jgroups.protocols.TP$IncomingPacketHandler.run(TP.java:1300)
            at java.lang.Thread.run(Thread.java:619)
    
  • Heap retention due to multiple org.jboss.remoting.transport.socket.ServerThread objects.

Environment

  • JBoss Enterprise Application Platform (EAP) 4.3 CP02.
  • 4 clustered instances of the application server running on two host machines.

Subscriber exclusive content

A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.

Current Customers and Partners

Log in for full access

Log In

New to Red Hat?

Learn more about Red Hat subscriptions

Using a Red Hat product through a public cloud?

How to access this content